Senior Computer Vision Engineer (Scientific Imaging)

Itransition·Узбекистан·14 сент.

Senior Computer Vision Engineer (Scientific Imaging)

Ориентир по рынку≈ 4 500 – 5 200 $наша приблизительная оценка по 17 вакансиям этой роли и грейда, у работодателя вилка не указана

🌍 УдалённоSeniorПолная занятостьАутсорс

≈ 4 500 – 5 200 $наша приблизительная оценка

Стоит ли откликаться

Стоит откликнуться, но спросите про деньги

Вместо вилки в объявлении «Competitive compensation»: что это значит в цифрах, придётся выяснять. Вакансия живая, задачи расписаны.

3 из 4 в норме

Зарплата не указана: ваша цель, нижняя граница и фраза для первого созвона

Деньги

Вместо вилки в объявлении «Competitive compensation». Ориентир на созвон: столько платят в 17 похожих вакансиях ML Engineer уровня Senior.

платят меньше4 500 – 5 200 $ у половины таких вакансийплатят больше

Вилки нет

Живость

Висит 37 дней, признаков фиктивного найма нет.

Живая

Описание

Расписано подробно: обязанностей 11, требований 14. Картина ясна до первого созвона.

Обещают: ДМС, гибкий график, уроки английского и компенсация спорта
Известно: в работе ИИ-инструменты для кода
Не сказано: как оформят и из каких стран можно работать

Подробно

Что требуют

ОбязательноComputer VisionPythonNumPy

Надбавка к зарплате на этом уровне: Python +14%.

Ясно

Полный разбор в Pro

Это короткий разбор. В Pro полный: под эту вакансию и ваше резюме

Сколько просить

Не продешевите: сумма под ваше резюме, нижняя граница и готовые фразы для торга

Ваше резюме против требований

Пройдёт ли ваше резюме фильтр: что совпало с требованиями и что дописать до отклика

Спросить у работодателя

10 вопросов именно к этой вакансии: что спросить на первом созвоне

Когда откликаться

Стоит ли спешить: за сколько дней Itransition обычно закрывает вакансии

30 дней Pro: полный разбор каждой вакансииЧто ещё входит в Pro

Вакансия на английскомПереведёт заголовок и описание вакансии на русский

Чем предстоит заниматься

Assess the first images: what is visible, at what contrast, where the method will and will not work

Define image-quality requirements the acquisition partner must meet, and check every delivery against them

Build preprocessing: tiling, thickness normalisation, denoising, seam handling

Build the conductor tracing: linear-structure filters, trace graph, branches, rejection of look-alike structures

Run the feasibility gate for the second defect type on a simulator and on reference samples with known defects; the outcome may be a documented no-go

Own the method: the physical and algorithmic hypothesis, pipeline architecture, thresholds and the policy for uncertain cases, validation design, the technical decision on every miss, and technical sign-off

Review the second engineer's implementation and keep the end-to-end pipeline reproducible

Work with the acquisition partner's engineers, together with the technical lead, on image format, calibration, metadata and stitching problems

Tune the pipeline on the first images from an installed blade and report what changed and why

Work with the NDT expert and the blade engineer on thresholds and on every miss

Be on site at one test campaign at a wind farm (about one week) to check images before the equipment leaves

Наши требования

4+ years in computer vision or image processing on real, imperfect images

Classical image processing at a working level: filtering, ridge and line detection, morphology, segmentation, connected components, graph-based tracing

Experience with images too large for memory: tiling, overlap, reconstructing coordinates in the full image

Strong Python with NumPy, OpenCV, scikit-image or equivalents

At least one project where the data was scarce and you had to design how it was labelled and validated

Experience quantifying image quality (contrast-to-noise, resolution, blur, repeatability) and designing controlled experiments that separate an imaging limitation from an algorithm limitation; any modality

At least one project where you moved into a modality you did not know before (medical, microscopy, satellite, industrial, scientific) and can explain how you got up to speed

English good enough to read papers and talk to the client's engineers

Radiography, CT, NDT, composites, weld or semiconductor inspection

Physics or engineering background: attenuation, scatter, beam hardening, geometric unsharpness, detector correction

Simulation of imaging (any X-ray or optical simulator)

Small-dataset deep learning: segmentation, transfer learning, hard-negative mining

Experience influencing how data is physically acquired, not only processing what you were given

Use of AI coding agents for prototyping and tooling

Мы предлагаем

Projects for such clients as PayPal, Wargaming, Xerox, Philips, Adidas and Toyota

Competitive compensation that depends on your qualification and skills

Career development system with clear skill qualifications

Flexible working hours aligned to your schedule

Options to work remotely

Corporate medical insurance covering services of private and public medical centers

English courses online

Corporate parties and events for employees and their children

Internal conferences, workshops and meetups for learning and experience sharing

Gym membership compensation

5 days of paid sick leave per year with no obligation to submit a sick-leave certificate

Дополнительно

We are looking for a Senior Computer Vision Engineer to lead the two-person core computer-vision team of an X-ray inspection proof of concept for wind turbine blades: six to eight months with a path to a product. Around you: a technical lead, a project manager, shared QA, an NDT expert and a blade engineer part-time, and the acquisition partner's engineers

The input is a single stitched X-ray image of a whole blade, tens of thousands of pixels long, delivered by the acquisition partner. Reconstruction from raw detector frames is outside the role and outside the initial scope. The first task is to find a thin, dense metal conductor inside the composite, follow it from one end of the blade to the other, and tell where it is broken. The second task is a feasibility question: evaluate whether missing adhesive or an open separation in a joint deep inside the structure produces a detectable and repeatable indication, and at what size. We expect classical image processing to be the starting point: normalising brightness across material that changes thickness by an order of magnitude, handling stitching seams, tracing a linear structure through clutter, calibrating a threshold on physical reference samples. Learned components are used only where experiments show they add value

You do not need X-ray or non-destructive-testing experience. We teach the physics and pay for the reading time. What we need is someone who has entered an unfamiliar imaging domain before and got to a working method fast, and who can tell the difference between "the algorithm missed it" and "there is nothing in the image to find". One honest outcome of the first phase is that a defect type cannot be seen reliably; we want the person who can show why